Riemann Hypothesis 4

Bernhard Riemann
12 followers
Part four of talk on the Riemann hypothesis. I
Introduced the idea of Fourier
Analysis, which is to take an arbitrary
graph or signal or data basically and express it in as a sum of
What's called what are called sine waves the simplest kind of up-and-down wave oscillation?
trouble is that
Fourier
because these sine waves are
functions that
Never get very big or very small
They don't go high or low. They're like this
They're much easier to
Create a signal of the same type
Then create a signal for example like this kind of thing if I wanted a function that goes up up up up up forever
It's very very hard
It's not impossible to make that with the Fourier idea, but it's very hard to figure out how to do that with the Fourier idea
and
Laplace
Had a variation of the Fourier idea
I'm not sure if you thought of it that way at that start and it involved what are called exponential functions
exponential functions are like like when you say exponential growth like
They're the way compound interest grows or bacteria in a dish if they had given enough food or sometimes population grows
Exponentially they can grow somewhat fast really fast kind of moderately fast
There's various variations on an exponential growth function, but they all grow as opposed to oscillate and wiggle and
it turns out that
somewhat like the Fourier idea of
Figuring out what kinds of wiggliness do you need to make up like this white square wave function you can think
If I've got an arbitrary function
Something that's done not one of these standard examples of exponential growth
Can I compare it to these guys and get an idea of of how to analyze the function?
So the idea is to describe a function by comparing it to various rates of growth. I'm gonna leave it vague at that point
And it turns out the idea actually goes back to Euler, but Laplace has his name on it because he did a lot with it
Turns out these aren't just similar ideas. They're really the same idea
And I want to explain a little bit about that because it brings up another wonderful topic and
That is the connection between Fourier and Laplace is through complex numbers
So let me just give you a very brief tour of the idea of complex numbers
The geometry of ordinary real numbers. Well, what do I mean by the geometry of numbers? I
Mean that if I have a number line just the ordinary real number line
Here's zero. Here's one two, three four
Minus one minus two minus three. I should mention
If I'm gonna discuss complex numbers, those can seem very scary kind of a weird non real kind of number
Not too long ago a few hundred years ago negative numbers were thought of as spooky and non real and you should never mention them
But we know better now we all get into debt. We understand negative numbers pretty quickly, hopefully
What I want to meet what I mean by the geometry of real numbers is that
One of the things that real numbers do is it encodes?
scaling
That if I look at the number line if I multiply
Any number like n or let's say X and I take 2 times X
What does that do it takes 0?
to 0
It takes 1
To 2 it takes 2 to 4. It takes 3 to 6
It takes minus 1 to minus 2 minus 2 to minus 4 it
stretches or
Scales it doubles the scales
The scale of things it's zoom. It's sort of like a zooming in
We've see that a lot in sort of computer applications. You can zoom in and out of things like Google Earth or something
But it's a very basic idea of scaling up
By a certain amount what about multiplying by a negative number that has a rather interesting idea minus 1 times a number
That takes the original number line
Well 0 still goes to 0 but 1 goes to minus 1
Minus 1 goes to 1 2 goes to minus 2
Minus 2 goes to 2 it flips it
It flips the direction or another way to say it
Is that minus 1 multiplication?
Corresponds to 180 degree rotation
It's the only kind of rotation that makes sense on a line
You can't have a line if you're living entirely on this number line. You can't couldn't do a 90 degree rotation
It wouldn't make any sense
Although we'll see pretty soon. We will let make it make sense for numbers
What about if I do this scale let's go back to the doubling scaling operation
What if I did that repeatedly what if I
took the number line and
I
Doubled everything. So let me just try I'm not gonna track the progress of everybody
I'm gonna track the progress of one if I double it I go to two. I
Double it again. It goes to four. I double it again
It goes to eight and of course everybody else is getting doubled at the same time
Let's just think about where one goes. So it goes one two four or that's two squared and
Then eight is two cubed and then sixteen is two to the fourth
Repeated scaling is
exponent exponents and
so
Exponential growth. Remember I was saying that the Laplace's idea was to take these very
basic kinds of growth functions
Exponential growth and use those to describe any function. Well
Exponential growth is just repeated scaling
It's just repeated multiplication by the same number
It turns out that our old friend E and I mentioned a while ago
It turns out that's actually a better one to use than two, but we can think about as one two, four, eight sixteen
somehow and again, I'm just trying to give it a sense of the
vague a vague sense of the connections here is
that there's a way to think of these exponential growth functions that Laplace was so interested in using as building blocks and
Think of them as this is what happens when you repeatedly use the geometry of real numbers
that what real numbers do a big role they have is
Scaling up and scaling down
Multiplying to get bigger divide and you get smaller and if necessary flip to negative a very limited kind of rotation
well
Complex numbers. Let me give you a very brief definition of complex numbers
Probably heard of them a little bit
You introduce this a magic number called an imaginary number called I and
With the amazing property that I squared is minus one or in other words, it's the square root of minus one
This is not a number we had before
There was there is no real number nor ordinary kind of number of squares to my negative numbers much less minus one
not only do we introduce I we allow ourselves to multiply to make five times I or minus seven times I and
We even allow ourselves to add those to ordinary real numbers
So it's things like 3 plus 2 I minus 7 plus 5 I etc at this level
It's what mathematicians love and what almost everybody else gets very very confused by or or disturbed by it's purely formal
I'm just introducing new symbols and
Trusting that they'll be useful. Well, you don't just introduce the symbols
You have to introduce the rules for manipulating the symbols still that seems a little dry, but we'll see why we do it in a second
How would you multiply these guys?
Well, what you want to do is you won't don't want to totally reinvent the rules of algebra
We'd like for the distributive law to hold this idea that I talked about before that if I have two things
Some times the sum I'm gonna multiply this times this this times this this times this this times this and so that's written out here
Okay. Well three times minus seven. That's not hard. It's minus 21
Three times five I well, hopefully the eyes not doing anything weird here and I just say well if I have five
Eyes, whatever those are like five oranges and then I have three groups of five oranges. I'm gonna have 15 oranges
Well, okay. So let's write that as 15. I
Do I times minus 7?
Well, hopefully the usual rules of algebra hold it'll be minus 14 on and then here 2i times 5i
well, let's group the real parts together 2 times 5 is 10 I
Times I is I squared and now here's where the special nature of I comes together. Ah, I squared that's supposed to be minus 1
That's supposed to be the new feature of this guy
So I don't have to leave that as it is
I translate that into a minus 10 and now these guys are just ordinary real numbers
I'm gonna group those together get minus 31 and
These guys I got 15 of something and then minus 14 of the same kind of thing
Whatever that thing is that should amount to just one of those things 15 I minus 14
I should be on and so there's an example of multiplying complex numbers and any multiplication of complex numbers is very similar
You can you can now do it if you want for any kind of numbers
But what does that do geometrically?
Okay. Well
What we're gonna do is we are going to put this into a geometrical form and
we're going to talk about what's called the complex plane and
so
Instead of just having a number line
with 0 and 1 and 2 and minus 1 and minus 2
We are going to have we're gonna put I I
claim it makes sense to put I up here and
minus I down here and
To I here and minus 2 I here and then what about all those combinations?
Well, let's see 2 plus I we're gonna do is we're gonna say if it's 2 real parts and 1 imaginary part
We're gonna put it over here
That's 2 reals and 1 imaginaries. We're gonna call that 2 plus I and
So for example 3 minus 2 I
Would be right here
Okay, so that's a kind of a funky way to introduce geometry into this. Does it make sense? Is it useful it you bet it is
what I'm gonna do is instead of
Repeatedly multiplying by 2 like I was talking about with the number line and talking about the geometry of real numbers
I'm gonna repeatedly multiply by I
So I times 1 I'm gonna start with 1 and I'm gonna multiply by I that's gonna be I that's up here
I times I
Okay, so that's another factor of I well that by definition. That's minus 1 I
Times minus 1 I'm gonna melt another multiple of I oh, that's gonna be minus I and
I times minus I that's a tricky one, but let's see I times I is I squared is minus 1
Another minus you get back to 1 so let's see what happened
I started with 1 and I repeatedly multiplied by I every time I rotated 90 degrees and got back to where I want
I started after 4 times
What about I times 2 plus I okay? I times 2 let's look at the algebra here, and then we'll look at the picture I
times 2 plus I is 2 I
Plus I squared I times I oh I squared is my another name for minus 1 minus 1 plus 2 I where is that?
That is minus 1 plus 2 I it's here if I look at
This and this that's a 90 degree rotation these guys are the same distance from the origin
But I've rotated by 90 degrees so again
Multiplying by I was a 90 degree rotation one more example if I start out with minus 1 minus. I that's down here and
I multiply it by I just using these formal rules of algebra I get minus I
Minus I squared but minus I squared cleverly enough was plus 1 I get 1 minus I
that's 1 minus I and
again
It's a 90 degree rotation, and you can show that this always works
that multiplication by I is always a 90 degree rotation and
So complex algebra encodes rotation not just 180 but in fact and not just 90 in fact
you can encode any kind of rotation you want with complex numbers and
So the algebra of complex numbers just as one simple rule I squared is minus 1 it turns out that you can leverage that
into a way
to encode any kind of rotation as well as scaling that's a much more sophisticated kind of geometry and
It still is just coming from root water really very simple algebraic rules unfamiliar and a little bit a tiny bit unfamiliar
But really simple algebraic rules
so remember that if I repeatedly multiply by 2
And I look at powers with real numbers you get exponential functions
turns out that when I repeatedly multiply with I
when I do
When I do the same kind of repeatedly multiplication of
Putting I in an exponent for example, then it turns out I get rotation. This is a utterly
Incredible formula due to Euler that I don't want to
explain completely, but it says
cosines and sines
Which have to do with oscillation remember cosine and sine where these kinds of waves the red waves?
Those have to do with oscillation
Oscillation has a deep connection with rotating in a circle
If I rotate it if I look at something rotating in a circle for example all right
Let's write let's draw it here. I'm running out of my paper. Okay here
If I draw something rotating in a circle
I'm just gonna look at this rotating. I'm gonna see if you can see my finger, okay
Pay attention only to how high it is not to how far left or right it is that's going up down up down
It turns out that that
if I just traced out the up-down behavior, it would exactly be a sine wave and
so rotation
Has to do with these fancy waves or these very nice ways called a sine wave and
It turns out that the way to express that in a formula
Again, this is one of those formulas that I figured I had to write down even though
I didn't want to explain completely is that the two fundamental sine wave
Pictures which are called sine and cosine you put those together with with complex numbers
It turns out that's the same as repeated rotation
An example of that is probably even an even more famous formula is that e to the I pi equals minus 1
Combine some of the absolute most important numbers in mathematics e which we've got a hint of its importance before pie
Which has to do with circles I which we now see has something to do with rotation
It says what this says geometrically. I know it's not obvious. It says that rotating by half a circle
Turns out that pi exactly gives you half a circles worth of rotation
It's the same as flipping minus 1 so I want to understand. This is a geometric
Statement, it's just the statement that a very special case of rotation is
That if I rotate I rotate I rotate I rotate I rotate if I do exactly
180 degrees of rotation
Which it turns out corresponds to pi worth of rotation in a certain system that
corresponds to just taking like
the number 1 and turning it into minus 1 or the number 2 plus 2 and turning it into minus 2
Very famous formula, and it has a geometric interpretation
So if you ever see this I know I haven't given you any of the details
But if you ever see this formula you want to maybe remember it looks like an algebra formula
But the way to really really make sense of it is that actually there's a geometrical meaning
It says rotating by a half circle is the same as multiplying by minus 1
So what I want to get out of this is simply again one of these vague connections that
Complex algebra and especially doing exponents with complex numbers
turns out to be that encodes rotation and
Especially it encodes like sort of repeating rotations over and over again
And so what does that have to do with the plus and Fourier story it turns out that?
As I said the kinds of functions we wanted to
We wanted to model let me go back to those the kinds of functions. We wanted to model were like this
They were like this the chebyshev psi function
That is not a function that looks like it's
Oscillating in any way well, maybe the little bumps up and down above, but the fundamental thing that this function is doing is it's growing
That's not something that looks like
This and
So you wouldn't wouldn't think that wave behavior would actually make much sense to describe that
Laplace comes along and says oh, that's growth
I can understand growing functions by comparing them to standard examples of growing functions
Called exponential functions things like 1 2 4 8 16
Fourier still it's still though
Fourier is an incredibly powerful theory and in some ways
It's prettier than the Laplace theory and the beautiful thing is if you bring complex numbers into it you can pretend
They're the basically the same theory
If you look at Laplace and you allow complex numbers into the picture it turns out it becomes Fourier or really there
It's a synthesis of the two called Fourier and Laplace
So amazingly enough there's a way to
understand functions that grow
By sort of pretending that they're really oscillating and that's involves putting I into this putting complex numbers
So here's the the idea so far as we have it
You start with
some function F
Some sort of picture of growth like for example the the prime counting functions you apply this amazing Fourier Laplace idea
Which is substantial to learn how to do, but you can you can do it
what you get is something that
Gives you a map of the behavior of this function in the complex plane. Let me see if I can have some paper
Here's the complex plane and
It gives me a map with certain special locations. Maybe 2i is special or maybe
2 plus i is special or maybe
0.75 minus
0.9i
Turns out that it identifies certain very special points in the complex plane
Based on the kinds of growth and oscillation of F. It's a very weird and clever way to take a function and
Identify how fast is it growing? Is it a combination of different kinds of growth?
Does it also have some oscillation at the same time some wiggles? How fast are the wiggles?
How big are the wiggles?
That kind of stuff is encoded in a very clever way in a map of the complex plane just various dots
Basically on the complex plane these special points are called poles and zeros in the complex plane
And it then turns out another part of it that I'm not going to explain. I don't think at all is
that
You can bring you can combine two other ideas that don't seem to be related complex numbers and calculus. I
told you how complex numbers come from algebra and
Then told you a little bit about how they really encode geometry
Well, it turns out combining them with calculus is an amazingly powerful thing to do
So you can do that with any function and it's a very useful thing to do
Could we do it for these functions like the prime counter or the modified the Chebyshev prime counter?
That this is this story
that I've been sort of trying to work up to in the last couple parts is a lot of why the Riemann hypothesis actually read a
Description of what the Riemann hypothesis technically says it doesn't seem to say anything to do with prime numbers
It's because of this rather intricate chain of ideas that we start out with something very simple counting primes
this stair step kind of function that we see
here a prime counter and
Instead of talking about the prime counting function the Riemann hypothesis seems to say something about this weird thing about the complex plane and some
Function and poles and zeros. What the heck does that have to do with anything?
It's because of this kind of story and these amazing combination connections between growth and oscillation
scaling and rotation
and
then
The question is maybe this is a totally useless thing to do. Maybe we're taking something
That's already hard to understand like the prime counting and just making it into an even more complicated guy
Yes, we have some techniques from combining complex numbers in calculus
But maybe it's just a bad idea turns out and I'll do this in the next
next
Section turns out there's an amazing calculated coming a connection. That's the word between Laplace
that Laplace idea of how to understand things in terms of their growth and the zeta function and
We already knew that that had some pretty special connections with primes and we'll do that in the next part
